Wellbeinn

I worked alongside the CEO and co-founders to shape product strategy, not just execute on it.

• Turned the founders' ecosystem vision into an architecture. The founders saw the band as the entry point to a multi-device ecosystem (band, ring, smart scale, recovery gear). My job was making that buildable: a data model and a BLE abstraction designed so every device after the first integrates with no re-architecture. One device exists today; the architecture is ready for the rest.
• Differentiation thesis. My strongest push: differentiate on what sleep and rest can do for the user, and anchor every metric in current science instead of shipping metrics for their own sake. It is now the company's stated positioning.
• AI roadmap. I designed the AI track that becomes the next post-MVP differentiator: proactive conversational agents that initiate based on user state, action recommendations driven by real-time wearable data, and an aggregator that reconciles them across devices, with the hiring it implies already in the team plan.

Choosing the manufacturer was the highest-stakes decision on the table: pick wrong and the company redesigns its product around a dead device. I evaluated three manufacturers against a Go / No-Go framework I wrote for the call. Two failed it on technical grounds. The third became the production partner only after a full audit. For that device I built the BLE protocol stack myself rather than rely on the vendor's documentation gaps, and filed formal bug reports as I found them.

I built the Flutter and Firebase platform alone, off a blank repository: layered architecture, an end-to-end test suite, and releases shipped to the App Store and Google Play. I could work alone because I built the engineering system around myself first: a development methodology of agentic loop orchestration with self-verification gates, so one person could hold the quality bar of a funded team. Over five months that came to 1,300+ merged pull requests and 12,000 automated tests holding 84% line coverage. Post-launch operations came with it: bug triage, support response templates, and backlog stewardship, so a small crew can take on real users without the support load swamping them.

The edge isn't the hardware. The hardware is a commodity. The edge lives in the algorithmic and scientific layer on top. I reviewed more than 100 peer-reviewed papers so the metric design rests on evidence, and stood up a 17-metric foundation the next waves keep extending. The discipline cut both ways: metrics the evidence couldn't support got pulled before they shipped.

The founders' vision was always more than one device, and the call was to pay for that future up front: the lead device ships now, and every wearable added later drops into the existing engine with no rebuild. The most important technical bet I made, and the foundation for where the company's defensibility ends up living.

The temptation in consumer health-tech is to license a third-party wellness scoring engine. I argued for proprietary metrics built on more than 100 papers of evidence, with the discipline to retract metrics when the evidence does not support them. That credibility gets the product a hearing with the people a health claim has to survive: a skeptical clinician, a cautious distribution partner, a user who has been burned by wellness scores before.

The long-term moat is something else: an AI agent that feeds on the user's own data across every device in the ecosystem and delivers a hyper-personalized experience back. Hardware is a commodity. Science gets you in the room. The agent fed by the user's own data is what keeps you defensible.

Quality and architecture choices that looked like overkill on a five-month MVP were really a bet that the company would need to scale past me. The proof arrived on schedule: the CTO I hired inherits a platform they can extend in week one, not a codebase that needs a quarter of cleanup before any new work starts.